José P.
Da Silva
*a,
Edgar V.
Bastos
a,
Luis F. V.
Ferreira
b and
Richard G.
Weiss
c
aFCT, Universidade do Algarve, Campus de Gambelas, 8005-139, Faro, Portugal. E-mail: jpsilva@ualg.pt; Fax: +351 289 800 066; Tel: +351 289 800 900, ext. 7644
bCentro de Química-Física Molecular, Instituto Superior Técnico, 1049-001, Lisboa, Portugal
cDepartment of Chemistry, Georgetown University, Washington, DC 20057-1227, USA
First published on 31st October 2007
The photochemical behaviour of the herbicide napropamide is studied on cellulose and silica surfaces, using steady-state and laser-flash diffuse reflectance techniques. The results are used to probe how the reaction sites of the host matrices influence the photo-reactive pathways. Napropamide undergoes reaction when irradiated with UV (lamps) or visible (sunlight) radiation on both solid supports. The nature of the intermediates and final products depend strongly on the presence or absence of molecular oxygen. The triplet state of napropamide adsorbed on cellulose is detected by both time-resolved luminescence and transient absorption spectroscopies. The triplet sate was not observed on silica, but transients which include the participation of molecular oxygen are detected during flash photolysis studies. The keto intermediates of the photo-Claisen rearrangement products are observed on both solids. Substituted 1-naphthols from photo-Claisen reactions and 1-naphthol are among the main reaction products. 1,4-Naphthoquinone is a major photoproduct in the presence of molecular oxygen and is expected to be prevalent when napropamide undergoes photodegradation in the environment (i.e., after being applied to plants and fields).
![]() | ||
Fig. 1 Simplified scheme for photo-Claisen reactions of napropamide. |
Napropamide has been reported to undergo photo-Claisen reactions in aqueous media.2,3 Photo-Claisen reactions of 1-naphthyl alkyl ethers are known to proceed mainly from their excited singlet states via homolytic scission of the C–O bond between the naphthoxy and alkyl moieties (Fig. 1).4 The 1-naphthoxy and alkyl geminate radical pairs thus formed undergo cage recombination, leading to the Claisen products, after tautomerization of the initial keto intermediates (I11 and I22, Fig. 1). The fraction of radical pairs that undergoes cage-escape usually results in the formation of 1-naphthol if the medium contains abstractable H-atoms.5 Although H-atom abstractions are slightly endoergonic from most of the C–H bonds in a polymer such as cellulose, they can occur when the naphthoxy radicals live for a long time and because there exist many ‘defect’ sites in this complex polymer that afford more attractive C–H bonds as H-atom donors. Additionally, when molecules of napropamide are either aggregated or in high local concentration, bimolecular formation of naphthol, involving abstraction of an H-atom from one molecule by a naphthoxy fragment of another, can become important. Regardless, the fate of the radical pairs is sensitive to the nature of their immediate environment, and that sensitivity can be monitored by the distributions of photoproducts obtained.4 Thus, the excited states and other intermediates involved in the photo-Claisen rearrangements (and related photo-Fries rearrangements of aromatic esters) are very useful in assessing the influences of different reaction environments on molecular motions over short distances.4–11)
The specific photo-Claisen products from napropamide are N,N-diethyl-2-(1-hydroxynaphthalene-2-yl)propionamide (A) and N,N-diethyl-2-(4-hydroxynaphthalene-1-yl)propionamide (B), as well as the cage-escape product, 1-naphthol (C). Although the photoreactions of napropamide have been examined on soil surfaces,12,13 we are unaware of any spectroscopic or photoproduct distribution studies employing it at solid–gas interfaces.
Here, we examine both the nature of the transients and the distributions of the photoproducts from napropamide adsorbed on the solid surface of silica and at two locations on cellulose. This study provides mechanistic information about the interactions between napropamide and its intermediates and these solid surfaces, and insights into how napropamide and other systemic herbicides and pesticides with similar structures14 are degraded by solar radiation when they are applied to soils and plants.
Washed cellulose samples (to remove napropamide at external surfaces) were also prepared and analysed. Typically, 50 mg of sample was mixed with 5 mL of hexane and allowed to equilibrate during 1–2 min. The mixture was then filtered and the solvent was evaporated. This process resulted in the removal of 10–15% of the napropamide at 50 µmol g–1 initial loading.
After irradiation, a weighed amount (10–15 mg) of sample irradiated within a 1 cm2 area was placed in 1 mL of methanol, the mixture was shaken vigorously, and the solid was allowed to settle. The liquid was then analyzed by HPLC.
The amounts of napropamide and 1,4-naphthoquinone were determined by comparing the peak areas from injections of a known amount of the methanol extract with those from a calibration curve obtained with standard solutions. The concentrations of the other photoproducts detected at 300 nm were estimated by assuming that their extinction coefficients are that of 1-naphthol at this wavelength. The reported uncertainties are the standard deviations calculated for more than five measurements. The quantifications were performed only on samples irradiated in air equilibrated conditions. Analyses of silica and cellulose samples kept in the dark for more than three months showed no sign of napropamide degradation. Extracts of non-irradiated and irradiated samples were also analyzed by GC-MS.
Details about the equipment used for ground state absorbance and calculation of the ground state absorbance spectra, the HPLC and GC-MS instruments and the conditions of their operation, and the laser apparatus are described in the ESI.†
![]() | ||
Fig. 2 Remission function spectra at room temperature of non-irradiated napropamide on silica (1, 50 µmol g–1) and on cellulose (2, 2.5 µmol g–1) and irradiated in air on silica (3, 50 µmol g–1) for 1 min at 254 nm, 1 cm from the lamp surface. |
![]() | ||
Fig. 3 Emission spectra of napropamide on cellulose (50 µmol g–1) in air after pulsed laser excitation (337 nm, ∼1 mJ pulse–1, 600 ps pulse width, 500 ns gate width). Spectra were recorded at delays of 1.0, 3.0, 5.0, 7.0, 9.0, 11.0, 13.0, 15.0, 17.0, and 19.0 ms (from top to bottom) after the laser pulse. |
The triplet state of napropamide adsorbed on cellulose was also detected by transient absorption spectroscopy. The triplet–triplet absorption of similar naphthoxy derivatives has been reported to have a maximum near 430 nm.20–23 Naphthyl acetates show a somewhat lower absorption maximum at 417 nm.5 The transient absorption spectrum of napropamide on cellulose obtained at pulse end (∼20 ns) shows a sharp absorption band at 340 nm and the expected triplet–triplet absorption of napropamide centered around 425 nm (Fig. 4a). After 20 ms, the absorbance at 340 nm decreased while the one at 425 nm nearly vanished.
![]() | ||
Fig. 4 Transient absorption of napropamide (a) on cellulose (50 µmol g–1) in air-equilibrated conditions at ∼20 ns and 20 ms after a laser pulse, and (b) on silica (50 µmol g–1) under argon and oxygen atmospheres at ∼20 ns after the laser pulse (266 nm excitation, 6 ns FWHM, 30 mJ pulse–1, 500 ns gate width). The inset shows the transient absorption obtained for napropamide in and air-equilibrated aqueous solution at pulse end (O.D. ∼ 0.4, 266 nm, 30 mJ pulse–1, 500 ns gate width). |
By contrast, the characteristic triplet–triplet absorbance of the naphthoxy moiety was not detected on silica, even when the pulsed excitation was performed on a sample under argon atmosphere (Fig. 4b). Failure to observe the napropamide triplet state on silica (even under an argon atmosphere) indicates that factors besides the presence or absence of oxygen determine the stability of the triplet states and the efficiency of their emission.
The absorption maximum at 390 nm and a low intensity absorption band between 400 and 650 nm, expected for a 1-naphthoxy radical,24,25 were not apparent in transient absorption spectra from samples with cellulose or silica supports. The lifetime of the geminal radical pairs from lysis of napropamide may be too short to be detected after the ca. 20 ns of our laser pulses—1-naphthoxy/acetyl radical pairs combine in acetonitrile at room temperature in less than 1 ns26—or to the presence of other transients absorbing in the same spectral region.
For comparison purposes, the transient absorption of napropamide was also studied in aqueous solution (inset of Fig. 4b). As mentioned, formation of the photo-Claisen products involves recombination of the radical pair, leading initially to intermediates I11 and I22 (Fig. 1). The tautomerization of these keto intermediates to the isolated ‘enolic’ forms, A and B, in solution occurs over periods of microseconds to seconds, depending upon the availability of acid or base catalysts,11 The keto intermediates exhibit an absorption band near 320 nm.26 On this basis, we assign the spectral feature with a maximum at 330 nm to compounds I11 and I22. The negative absorption change at shorter wavelengths is due to depletion of napropamide. The absorptions between 320 and 360 nm observed on silica and cellulose must be due to the keto intermediates as well as photoproducts A and B, especially at longer times (See Figure S-1†). On the silica surface, the absorbance increases between 330 and 530 nm under air and oxygen atmospheres (Fig. 4b). It is ascribed to transients from reaction of naphthoxy radicals and molecular oxygen (vide infra).
Silica a | Cellulose a,b | |||||||
---|---|---|---|---|---|---|---|---|
Initial concentration/µmol g–1 | 48.9 ± 2.2 | 48.4 ± 2.5 | ||||||
Radiation | 254 nm | Sunlight | 254 nm | Sunlight | ||||
a One standard deviation. b Samples prepared using methanol. c 1–3 PM | ||||||||
Irradiation time | 5 min | 14 h | 0.5 hc | 1 day | 5 min | 14 h | 0.5 hc | 1 day |
Final concentration/µmol g–1 | 40.5 ± 2.3 | 18.6 ± 2.5 | 35.2 ± 1.8 | 2.8 ± 2.3 | 43.3 ± 2.5 | 22.5 ± 1.5 | 37.3 ± 2.2 | 21.9 ± 2.9 |
The distributions of photoproducts from napropamide were first studied in aqueous solutions. The major products detected, A, B, and naphthol (C) (Fig. 3), are consistent with previous reports2,3 and account for nearly 100% of the consumed napropamide after ca. 15% conversion. At higher conversions, the initial photoproducts undergo secondary reactions, leading to unidentified species, and the mass balance becomes poorer.
At very low photoconversions (<5%) on cellulose and silica, the photoproducts, A, B, naphthol (C) and naphthoquinone (D) (Fig. 5), account for 60–75% of the consumed napropamide after irradiation at 254 nm or in sunlight. At higher conversions, the mass balance becomes poorer, indicating (as in aqueous media) that the initial photoproducts undergo efficient secondary reactions. The presence of photoproducts covalently bonded to the cellulose matrix especially,16,17,27 may contribute to the lack of a good mass balance. However, the absorption spectra of the extracted samples after irradiation are <5% as intense at all wavelengths than before extraction.
![]() | ||
Fig. 5 HPLC chromatograms at 220 and 300 nm as the detector wavelengths of the extracted product mixture from an irradiated (254 nm) sample of napropamide on silica (50 µmol g–1). Several unidentified minor products are apparent. The conversion of napropamide was ∼15%; peaks a, b, and c are unidentified products. |
The keto intermediates in Fig. 1 are known to be photochemically labile.26 The slower is their tautomerism, the more likely they can absorb a photon and undergo a secondary photochemical reaction. In fact, α-cleavage of the keto intermediates in the photo-Fries and photo-Claisen rearrangements to form dienic ketenes has been documented by flash photolysis studies.28 Thus, stabilization of the keto intermediates of napropamide on cellulose or silica may increase the fraction of reaction that does not yield eventually one of the classic photo-Claisen products (and account for the decreases in the mass balances with increasing conversion). HPLC analyses at 220 nm (Fig. 5) and other detection wavelengths of the extracts from napropamide irradiated on both solid supports indicate the formation of products containing only one aromatic ring (i.e., absorbing at 220 nm, but not at >300 nm; see Fig. S-2†). This observation is consistent with there being secondary photoreactions of the initially formed (major) photoproducts.
On cellulose, the photoproduct distributions are similar at napropamide loadings between 2.5 µmol g–1 and 250 µmol g–1. On silica, increased loadings result in higher relative yields of product c (see Fig. 5 and Fig. S-2†). A possible explanation for this observation is that formation of product c involves a bimolecular process. A photoproduct with a high retention volume, assigned to a dimer of product A2, was previously reported from irradiations of napropamide in aqueous solutions. In support of this contention, the photoproduct distributions are similar when irradiations are conducted on activated or unactivated silica.
Products A and B were identified by comparison with their reported spectra while the structure for C was deduced from comparison of its properties with those of an authentic sample. An oxidation product, 1,4-naphthoquinone (D), was observed after irradiations on silica in an air atmosphere. Its structure was established by comparison of its HPLC retention volume and MS spectrum with those of an authentic sample. Irradiations of cellulose samples in the presence of oxygen also lead to D, but in lower relative yields. D is not detected after irradiations under an argon atmosphere on both solid surfaces. The presence of D also contributes to the aforementioned increased absorption of the irradiated samples observed in spectrum 3 of Fig. 2.
Medium | Irradiation | Conversion | Photoproduct distribution | |||
---|---|---|---|---|---|---|
A | B b , c | C b , c | D b , c | |||
a Samples prepared using methanol. b Relative to product A. c One standard deviation. | ||||||
Water | 254 nm | ∼15% | 1.00 | 0.45 ± 0.05 | 0.10 ± 0.05 | — |
Cellulose a | 254 nm | <5% | 1.00 | 0.60 ± 0.10 | 0.40 ± 0.10 | 0.50 ± 0.15 |
Cellulose a | >290 nm | <5% | 1.00 | 0.80 ± 0.10 | 0.35 ± 0.10 | 1.30 ± 0.15 |
Silica | 254 nm | <5% | 1.00 | 0.75 ± 0.10 | 0.10 ± 0.05 | 0.90 ± 0.15 |
Silica | >290 nm | <5% | 1.00 | 0.60 ± 0.10 | 0.10 ± 0.05 | 1.95 ± 0.20 |
Similar behaviour was found on silica. The results suggest that a geminal radical pair on this solid support is not impeded from undergoing the rotational and short-distance translational motions that lead to formation of the keto tautomers of products A and B, but as with cellulose, processes that depend on translational diffusion of the napropamide fragments over long distances should not be important. The higher local mobility expected at sites on silica is also consistent with the failure to detect the triplet state of napropamide molecules adsorbed on its surface even under an argon atmosphere; deactivation processes can compete efficiently with intersystem crossing by napropamide singlets and subsequent emission from the triplets. The formation of product c only on silica is also in agreement with the possibility of additional reaction channels for the transients on silica surfaces.
The formation of 1,4-naphthoquinone from irradiations on cellulose in air, although in lower yield than on silica, was somewhat unexpected based on the time-resolved luminescence spectra. However, this result is more reasonable if one considers that some napropamide molecules reside at or very near the surfaces of cellulose matrices and others are buried within the matrix. Napropamide molecules at sites near the surfaces are much more exposed to molecular oxygen, and their intermediates can lead to formation of 1,4-naphthoquinone. Based on Beer’s law, electronic excitation (and, therefore, reaction) of napropamide molecules in these sites is more probable than in deeper, more constrained sites. Thus, the relative yield of 1,4-naphthoquinone on cellulose is expected to be (and is) highest at the lowest conversions: at ∼5% conversion, the A/D product ratio is one order of magnitude lower than at 25% conversion. To determine the role of non-entrapped napropamide molecules on the formation of product D, we irradiated a cellulose sample whose surfaces had been washed with hexane, a solvent which is unable to penetrate into the interior of bulk cellulose samples,20 but is able to remove surface-adsorbed napropamide molecules. As expected, only traces of product D were detected after irradiation of these samples.
Based upon these observations, the nature of the initial photoproducts from herbicides and pesticides is expected to be strongly dependent on their ability (as well as that of their intermediates) to diffuse on or to an air–solid surface (such as soil or a leaf). The diffusion of systemic herbicides and pesticides (structurally related to napropamide) to locations below plant surfaces should decrease the ability of molecular oxygen to come into contact with 1-naphthoxy radical intermediates before they follow a different reaction course and, thereby, decrease the yield of 1,4-naphthoquinone.
The formation of 1,4-naphthoquinone was also observed when 1-methoxynaphthalene was irradiated on a silica surface;30 it was attributed to reaction of a ground-state molecule with singlet oxygen. Radicals, such as 1-naphthoxy, are also known to be very reactive towards molecular oxygen.31 It is, therefore, likely that the formation of 1,4-naphthoquinone results from reaction of molecular oxygen with the naphthoxy radical or with 1-naphthol itself.32 The transformation of the oxygen adducts to yield 1,4-naphthoquinone can occur via several plausible routes.32
These results provide valuable insights into how napropamide and other α-naphtoxy herbicides and pesticides are transformed by sunlight after their deposition on natural surfaces. Based upon our observations, the diffusion of systemic herbicides and pesticides into the interior regions of plants after their deposition on surfaces lowers the ability of molecular oxygen to come into contact with 1-naphthoxy radical intermediates and, therefore, should decrease the amount of the initial material which is converted to 1,4-naphthoquinone. To confirm this hypothesis, other model surfaces need to be explored. Thus, we intend to investigate the photochemical behaviour of napropamide and its intermediates on ca. surfaces with cationic centers,9 since they are more similar than silica to the environments found in soil.
Footnote |
† Electronic supplementary information (ESI) available: Experimental details and product spectra. See DOI: 10.1039/b713369c |
This journal is © The Royal Society of Chemistry and Owner Societies 2008 |